Introducing our flexible forecasting platform for Zero Touch Networking and Digital Twinning, developed as part of the CLEVER Project 🔍
In today’s dynamic 5G and beyond environments, reactivity is not enough. Network automation must evolve into proactivity — where systems anticipate failures, forecast traffic, and autonomously optimize performance.
📶 Our platform uses real-time and historical data to forecast key metrics like:
- Channel Quality Indicator (CQI)
- CPU and memory utilization for Digital Twins
And more! ✨
📈 Tested with LSTM, GRU, and CNN models, the platform proved highly accurate in real-world use cases:
✅ CQI prediction with minimal error — enabling proactive handovers
✅ Resource-aware auto-scaling for robotic Digital Twins — guaranteeing SLA compliance
🔧 How it works:
- Kafka-based data ingestion & REST APIs
- Dynamic model selection & automated training
- Seamless integration with orchestration tools
- Designed for modularity and multi-metric forecasting
📊 See it in action below:
Forecasting CPU and Memory in Digital Twin Scenario
This is just the beginning. Our vision includes multi-stream data ingestion and online model optimization via NAS — bringing real AI into Zero-Touch Networks.
🔗 Explore the CLEVER Project: https://www.cleverproject.eu
Check the paper for more details! : https://zenodo.org/records/11032637
#CLEVERProject #DigitalTwin #ZeroTouchNetworking #AI #MachineLearning #5G #Forecasting #SDN #CQI #Automation #NetworkIntelligence #FutureNetworks #DeepLearning #EdgeAI #NeuralNetworks #TechInnovation #CLEVER #KafkaAI